WHAT IS THE BEST METHOD OF ALIGNMENT OF PUMPS HANDLING HOT FLUID DURING NORMAL OPERATIOH , SUCH AS BOILER FEED PUMPS. THE PUMPS ARE TO BE ALIGNED IN COLD CONDITION. HOW TO DECIDE REGARDING COMPENSATIONS FOR THERMAL GROWTH SINCE IT MAY BE DIFFERENT FOR SAME MAKE, MODEL AND RATING PUMPS.

Generally the OEM will supply expected thermal growths based on pumps at a given temperature and not sitting in the sun or severe environmental concerns. You must set valid 'cold' or initial align using correct procedures and valid tolerances. What is your coupling. Considering all factors, you'll not want ever to exceed an angle that removes over 80% of the diameterical clearance of the bearing. If you're not versed in performing alignments you may want to consider contracting out.

Set to OEM specs as a begining if you don't have valid hot data. No data: you may use Acculign hot data acqusition system to gather cold - hot and PM of running machines on a regular basis. This is a very good system basically approaching the accuracy of optics or as good. The across-the-coupling method should be avoided on large turbo-machinery and/or pedestal bearings. The use of lasers to setup prism and shoot around machines is very difficult and not a normal portable solution. The Acculign is one of the best and most cost effective solutions out there but not the only one.

Aligning machines are more or less a guesswork, especially in applications like the one you describe, even if you have thermal growth data from the machine manufacturer. Normally you only get data for vertical growth without considering pipe strain or other factors, and you have no data for horizontal movement. Ideally you would align the machine during running conditions, but naturally this is impossible. So how can we determine exactly how much the individual machine is moving, both vertically and horizontally, from off-line to running conditions?
The solution is found with Fixturlaser and their OL2R fixture. This fixture works together with the laser alignment system and is used to determine the actual movement from running conditions to off-line. Take a look at the Fixturlaser OL2R website. It is a patented solution making a big difference in howto solve this problem.

Good luck
Hans

I always went through two stages of alignment.
First, go with the OEM alignment figures.
Before start-up put two dial indicators on each of the bearing housings, one to measure vertical growth and the other to measure sideways movement.
Run the equipment up to full operating conditions.
Check your readings. Hopefully they should be acceptable and you can just record the "actual" movement, but keep running. If not, you have the exact change of alignment to make the corrections needed.

Hans,

The across-the-coupling method is very, very poor on large turbo-machinery or pedestal bearing machines - did I stress very poor, it can give you false data? I would not use it for wanting reliable data on costly large machines.

The Acculign method or optics allows monitoring of machines (all bearings) both front and back movements, piping, bases, skids and horizontal, vertical & axial movements, etc.

However, the across-the-coupling will probably do on small rigid frame machines. In the '70's we were using Dynalign or Dodd bars for across-the-coupling method using invar trusses and eddy current probes. Over time I think the across-the-coupling has proven itself as stated.

It appears caveman is using water stands and dial indicators ensuring heavy supports and/or sand-bagged or bricks of lead, water cooled tubing and in a location free of vibration.

Another handy feature of Acculign is: You can monitor problem machines quarterly for a year or so or through seasons if outside and easy so your millwrights can handle the data acquisition function giving you a PM program if needed for time study.

Sam,
I have attempted growth and movement calculations on very large feed water pumps using a Rotalign Pro with the M3 brackets. I mounted the brackets and took measurements while the machine was up and running. A few days later they shut it down and I took the second set of readings. I'm not sure if someone moved the brackets or what but my data was bogus. Could you explain a little more about this Acculign method?
Thanks,
Pete

Not easily! Tooling balls or indexing benchmarks are put on the machine and base/foundation (fixed location). May be located in radial locations in plane of bearings or axial. Also, these benchmarks or tooling balls can be put on any item of interest such as piping. It is a good and very cost effective approach.

Mr. Jack Essinger was the inventer and an icon of the alignment discipline.

The data are easily plotted by hand quickly or there are programs he has to do the calculation for you. A good millwright can easily acquire this skill. Google search or write me directly. I do address this topic in my book.

The acculign is referred to before in this thread and Sam appears to have experienced its usefullness for PM:

http://maintenanceforums.com/eve/forums/a/tpc/f/1191034...811068562#7811068562

A tread related to alignment:
http://maintenanceforums.com/eve/forums/a/tpc/f/1191034...341071662#4341071662

What is your book, Sam?

Hello guys,

Could you all help define a temp limit for hot alignment? Say 100 degrees Celcius. I mean that above this temp we should do hot alignment.

Josh,

Machinery Alignment - back to basics ~330 pages.

yahoo,

On this thread we generally only give or provide limited information. Afterall we are not writing a book or your report and generally are only given or dealing with limited information but htat is not always the case. We all have understanding of certain limitations and we are all human.

If a tolerance is 8 mils for a given machine and its coupling then you are going to run better within those limit boundaries. Machines, couplings, etc.. may not all be the same for a given temperature. You may have a steel frame motor and SS pump or??????? Calculations must be made and the right coupling selected and you can go by machine centerline height and delta temp for the expansion constant; egro steel = 6.3 X 10 E-6. If you know the temps, you may be able to engineer the right coupling or if the coupling with the most forgiveness is not totally adequate to allow for all expansion, then have a desired offset that is within a number limit so that the machines grow to toleranc specification with a short period of time.

Sam

Cannot find a link to your book in internet. You should put the title under your name here.

Only found this article by you:http://www.reliabilityweb.com/articles/Vibration_TroubleShooting_Guide.pdf